摘要
本文利用中国家庭金融调查(CHFS)数据,通过建立双重差分模型(DID)研究2018年个税改革对于家庭单元在风险性金融投资决策的影响,并进一步利用Q-Learning算法建立资源型和非资源型省份家庭强化学习模型进行仿真分析。研究发现,个税改革可以优化家庭金融资产配置和增加家庭金融资产;降低税率会增加家庭金融风险规避意识;考虑地区发展因素,资源型省份下家庭的政策敏感性更强。本文丰富了个税改革效益的研究方法,对优化个税改革政策实施和推进地区经济发展布局提供参考。
This article uses data from the China Household Finance Survey(CHFS)to study the impact of the 2018 personal income tax reform on household investment decisions in risky finance by establishing a double difference model(DID).Furthermore,Q-Learning algorithm is used to establish reinforcement learning models for resource-based and non resource-based provincial households for simulation analysis.Research has found that individual income tax reform can optimize the allocation of household financial assets and increase household financial assets;Lowering tax rates will increase household awareness of financial risk avoidance;Considering regional development factors,households in resource-based provinces have stronger policy sensitivity.This article enriches the research methods on the benefits of personal income tax reform,providing reference for optimizing the implementation of personal income tax reform policies and promoting regional economic development layout.
作者
高佳明
张丽
蒋梅
Gao Jiaming;Zhang Li;Jiang Mei(Xinjiang University of Finance and Economics,College of Information Management;Xinjiang University of Finance and Economics,Accounting College)
出处
《金融发展评论》
2024年第3期80-95,共16页
Financial Development Review
基金
新疆维吾尔自治区社科基金项目《拉动内循环经济背景下家庭金融行为及家庭金融风险防控研究》(项目编号:21BJY048)的资助
关键词
个人所得税改革
资源型省份
风险性金融资产投资
强化学习
Personal Income Tax Reform
Resource-based Provinces
Risky Financial Asset Investment
Reinforcement Learning
作者简介
通讯作者:高佳明,工学硕士,助教,新疆财经大学信息管理学院,Email:gaojiaming1997@163.com;张丽,管理学硕士,助教,新疆财经大学会计学院;蒋梅,经济学博士,讲师,新疆财经大学会计学院。
